Re: Balsa git version, SIGSEGV [not solved]



I was too quick.
It decided to crash during the lunch.

J-L

Le 10/06/2012 12:10:39, Jean-Luc Coulon (f5ibh) a écrit :
>Hi Peter,
>
>I've reverted the patch and no more crash... so far.
>
>Regards
>
>Jean-Luc
>
>Le 09/06/2012 20:17:35, Peter Bloomfield a écrit :
>>Hi Jean-Luc!
>>
>>On 06/09/2012 01:16:15 PM Sat, Jean-Luc Coulon (f5ibh) wrote:
>>> Hi,
>>> 
>>> I experience crashes of balsa.
>>> If I left it alone for a while, it disapears...
>>> 
>>> I've lauched it from gdb and you will find attached the backtrace.
>>
>>Thanks for the backtrace!
>>
>>It's not clear to me what's happening. Somehow a non-NULL mailbox in
>>the caller is becoming NULL in the called function. The only recent
>>change anywhere in that stack is 
>> 82f0d6822a2f7eded61986f18df6219bb2f3787d,
>>which added the g_warning when Balsa locks a subthread, resulting in
>>the warning at the top of your gdb log. I don't see how it could be
>>relevant, but you could revert that commit by reversing the attached
>>patch. Please let us know if it helps.
>>
>>The warning is useful in tracking down places where Balsa is 
>> preparing
>>to call some Gtk function from a subthread. That practice is now
>>frowned upon, and the thread-locking API will be deprecated and 
>> removed
>>some time soon. Many places have already been fixed; that code path 
>> for
>>filtering messages is one that has not yet had any attention.
>>
>>Best,
>>
>>Peter
>
>
>


Attachment: pgpYVI2fAUPaE.pgp
Description: PGP signature



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]